CFExpress Cards Speed Tests

Let me start off by stating, I do not expect that I will often use top end specs of an R3 or R1 camera. Heck, for the R5, I do use the 20 fps, but I do not fill the buffer, it is for a few second bursts.

That said, some speculation I have seen really has me wondering about the R1. The speculation is high FPS and high MPs. But, that equates to high data throughput. So, I ran some numbers starting with theoretical max write speeds of CFExpress cards (~1400 MB/sec of actual cards and 2 GB/sec for Type B). But, then I got to thinking no card ever performs as well as rated. So, I played with Bryan's data of # pictures before hitting the buffer on the R5. Based on the numbers, I get the R5 buffer being 4-6 GB, but the write speed of the ProGrade Cobalt card he used only being ~500 MB/sec, or 1/3 of its rated 1500 MB/sec.

This made me look online, and I did find this article. First, the difference among cards is impressive, even different capacities (Sandisk, looking at you). They published time to clear the buffer on the R5, not MB/sec, but they estimate the buffer to be 2 GB and 4-8 seconds to clear gives the write times to the cards would be 250-500 MB/sec. Way off of the spec. (BTW, they also got very different results to Bryan in terms of pictures to fill buffer, so, something is inconsistent between these tests).

Their write speeds using a card reader were 450-800 MB/sec. Again, way off the spec, but at least getting closer.

So, I have always known cards perform at less than their max spec. However, usually they are 50-75% of that spec, not 15 to 33%.

CFExpress, still better than SD UHS II (which are typically ~180-200 MB/sec), but they are not functioning nearly as well as is being widely assumed.

I assume with the R1, Canon will want to have something similar to the 1DXIII where it doesn't hit a buffer (rated at >1000 RAW images). But the 1DXIII is only moving ~600 MB/sec with its 20 MB sensor to do that. Canon can not go much higher in MPs before they run into limitations of the actual write speeds to CFExpress cards. Which means one element we may be waiting on a high MP/high frame rate camera is improvements to CFExpress cards. Specifically, looks like current CFexpress Type B cards are PCIe 3.0 x2 lanes and is currently maxed out at 2 GB/sec. Type C (similar in size to old CF cards) with PCIe 3.0 x4 lanes can do up to 4 GB/sec eventually and I saw reference to PCIe 4.0 is eventually coming. But, that also means, if you want the top end specs, we'll likely be looking at new cards.

Thanks for posting, Brant! I have typically used SanDisk cards but would probably have gone with the 128 GB capacity for the R3. I suspect I'll need pony up for a pair of the 512 GB SanDisks (at $600 each).

Thanks John.

The only other thing I would consider is heat. I have heard that there is a difference between card brands. This is a new review site to me, but, after a search, I found this, granted that was generated moving 85 GB of 8K video. In looking at the SanDisk review, they also found that SanDisk started throttling speed as it went above 65 C, but they still gave the 512 GB SanDisk their gold seal.

It seems that they are updating their comparison tables with each Type B card review. The Delkin Black is looking very good, and actually starting to get to much better "sustained" write speeds (double the Sandisk) when transferring while maintaining cooler temps. But, then looking at the data that they consider more real world, the two is virtually identical and 15 GB in 31 secs = 483 MB/sec. So, the main benefit may be cooler temps.

I use both the 128 GB (my original) and 512 GB SanDisk cards (bought for Laguna Seca trip). Both great and I haven't noticed a difference, but that is how I shoot. I have noticed that after pulling them out of the card reader, they are hot. No doubt about that.

Enjoy the R3 :cool:.

Thanks for the info, Brant. Like you, I tend to shoot in short bursts so probably a smaller card would be fine. Based on the file size of the R6 (slightly more MP than the R3), even a 64 GB will hold close to 3,000 RAW images. I can’t imaging shooting that many frames in a session, even at 30 fps, and I generally prefer more cards than bigger ones. If I shot video I may feel differently…but I don’t. So given that, I’ll probably pick up a pair of 64 GB CFe cards and see how it goes.

I also need to sort out the SD cards. I have a pair of 64 GB cards that I swap for my EOS R (they came in a bundle with the camera, along with a couple of video accessories that I gave away). However, they’re UHS-I cards and while that wasn’t an issue with the R, I think they’ll be too slow for the R3. I’ll pick up a pair of 64 GB UHS-II SD cards and see how they perform with writing RAW simultaneously to CFe and SD. If that slows it down, I’ll write just to the CFe and periodically Image Copy the files over to the SD during downtime (to mitigate against card failure).

I’ll definitely enjoy the R3, whenever it comes. In the meantime, I have an RF 70-200/2.8 showing up next week.

I too usually prefer multiple smaller cards.

I used some very inexpensive 64gb Sandisk SDII cards in the R5 at Laguna Seca and had no issues running into the buffer.

I did notice however a marked reduction in speed when downloading the images for processing and archiving.

As the owner of a dead SanDisk Extreme card, yes, multiple decently sized cards are the way to go. You don't want to switch cards too often, but you don't want to lose your only storage all at once either.

I've been using the 512 GB Sandisk CFexpress cards and so far the biggest burst that I've fired is when these two young grizzlies starting play fighting in front of me. I held the shutter button down for a solid 2 minutes, blasting 16 fps @ 20 MP the whole way through. I didn't check how hot the card got, but I'm happy to report that it did the job quite well.

https://photos.smugmug.com/Nature/Wi...44-Edit-XL.jpg

Quote:

---

Originally Posted by Joel Eade

I used some very inexpensive 64gb Sandisk SDII cards in the R5 at Laguna Seca and had no issues running into the buffer.

---

I had previously mothballed my UHS-I cards. After talking with you, watching you do this and looking at my own usage, they are back in circulation.

For my usage, UHS-I is fine, largely because the R5 has a surprisingly good sized buffer. Going with Bryan's data, you are still looking at 151 shots at 12 fps to UHS-I SD card. On the 5DIV, Bryan got 21-36 shots at 7 FPS. Huge improvement. I tend to shoot action shots 1-3 second bursts, which means, I could do ~5 x 30 frame bursts before the buffer filled while the R5 vs ~1 on the 5DIV. After that, it is a matter of how long to clear the buffer.

However, I bought the UHS-II and CFexpress cards. I am primarily using those, but the UHS-I cards in occasional use too and I have yet to have an issue. In the future, I'll probably still pick up CFexpress/UHS-II cards. Just nice to have faster transfer speeds and not having to worry about in camera limitations for when I finally see Big Foot in the woods. ;)

Quote:

---

Originally Posted by neuroanatomist

Thanks for the info, Brant. Like you, I tend to shoot in short bursts so probably a smaller card would be fine. Based on the file size of the R6 (slightly more MP than the R3), even a 64 GB will hold close to 3,000 RAW images. I can’t imaging shooting that many frames in a session, even at 30 fps, and I generally prefer more cards than bigger ones. If I shot video I may feel differently…but I don’t. So given that, I’ll probably pick up a pair of 64 GB CFe cards and see how it goes.

I also need to sort out the SD cards. I have a pair of 64 GB cards that I swap for my EOS R (they came in a bundle with the camera, along with a couple of video accessories that I gave away). However, they’re UHS-I cards and while that wasn’t an issue with the R, I think they’ll be too slow for the R3. I’ll pick up a pair of 64 GB UHS-II SD cards and see how they perform with writing RAW simultaneously to CFe and SD. If that slows it down, I’ll write just to the CFe and periodically Image Copy the files over to the SD during downtime (to mitigate against card failure).

I’ll definitely enjoy the R3, whenever it comes. In the meantime, I have an RF 70-200/2.8 showing up next week.

---

Regards to the RF 70-200 F/2.8, nice! Everything about it seems good, but the size and ability to pack it upright in particular makes it appealing. I am not sure what my first RF lens will be and when I will start making the transition, but that lens could easily end up in my bag.

After my last post, I took a look at the R3 specs a bit more. Canon has the buffer depth rated at 150 RAW images. Assuming 30 MB per file for the R3, 60 MB/sec UHS-I, 180 MB/sec UHS-II and 500 MB/sec CFE, you are looking at 2 fps, 6 fps and ~17 fps after the buffer has filled and 75 sec, 25 sec, 9 seconds to clear the buffer. If you ever fill it.

As Joel proved, UHS-I cards will really work just fine for most of us even in action instances. Most of us just will not take that many images that quickly. That said, UHS-II does seem to strike a nice balance.

Quote:

---

Originally Posted by Kayaker72

Regards to the RF 70-200 F/2.8, nice! Everything about it seems good, but the size and ability to pack it upright in particular makes it appealing. I am not sure what my first RF lens will be and when I will start making the transition, but that lens could easily end up in my bag.

---

My first RF lens was the 24-105/4L, which I bought along with my EOS R. That was before the 24-70/2.8L IS came out, and I now need to decide if I replace my EF 24-70/2.8 II with the RF version. For this sort of thing, I find analyzing the EXIF data from my library to be very helpful. As a Mac user, I used to use Aperture for my library organization, and I had an app called Aperture Inspector that graphed out usage by lens, focal length, aperture, etc. I was disappointed when Apple killed Aperture, and I was forced to use Photos for my library (I've been reluctant to move to LR even though I use PS CC – mostly because I prefer DxO PhotoLab for RAW conversions but it doesn't do well with library management). Photos is bearable. I found a Mac App Store download called PhotoStatistica that performs the same analysis and graphing functions as Aperture Inspector.

IMO, the f/2.8 vs. f/4 choice is more difficult with RF lenses. For a higher-end DSLR, an f/2.8 lens means better AF performance (more cross-type points, more high-precision points) and a brighter viewfinder, and those benefits apply regardless of aperture choice. For the R-series bodies, with DPAF and the EVF, the difference really just comes down to how much benefit one derives from the more light and greater OOF blur of a faster aperture. For the 70-200m focal length, it was almost a no-brainer since over 50% of my shots with the EF 70-200/2.8L IS II were taken wider than f/4.

For the EF 24-70/2.8 II, about 25% of my images are wider than f/4, which is still a pretty substantial fraction. But there the trade off is not only size/weight/cost, it's 70mm vs. 105mm at the long end. A bit deeper analysis shows that with the EF 24-70, about 20% of my images are taken at 70mm, but looking at the resulting image sizes only a total of about 6% of the images with the lens are cropped beyond the 1-2 MP loss from correcting rotation and minor framing adjustments, so most of the time a 70mm long end is fine.

But coming at it from the other direction, with a 24-105mm lens about 25% of my images are in the 70-105mm range (applies to the EF lens when I had it as well as the RF lens). So if I have the extra 35mm, I use them about as often as I use the wider aperture. Does that mean I need both? Probably I'll end up buying the RF 24-70/2.8 and seeing which I use the most. If I had to bet, I'll use the 24-70 more but take the 24-105 when I travel.

Quote:

---

Originally Posted by Kayaker72

After my last post, I took a look at the R3 specs a bit more. Canon has the buffer depth rated at 150 RAW images. Assuming 30 MB per file for the R3, 60 MB/sec UHS-I, 180 MB/sec UHS-II and 500 MB/sec CFE, you are looking at 2 fps, 6 fps and ~17 fps after the buffer has filled and 75 sec, 25 sec, 9 seconds to clear the buffer. If you ever fill it.

As Joel proved, UHS-I cards will really work just fine for most of us even in action instances. Most of us just will not take that many images that quickly. That said, UHS-II does seem to strike a nice balance.

---

I'd probably be fine with the UHS-I cards I have for my EOS R, but I already ordered a pair of 64 GB CFe and a pair of 64 GB SD UHS-II. :)

I went through a similar situation with a lens selection with an ultra wide vs wide angle for a trip to Mt Baker last weekend.

I was somewhat on a budget and because it was last minute I needed to find something either used or new locally.

I knew the 16mm wide end would be fine from my previous excursions up there. I knew the obvious practical choice was the 16-35mm f/4 IS vs the 16-35mm f/2.8 II or the 24mm f/1.4.

I really wanted the 24mm f/1.4 for other uses. And I knew up there I'd still get some great pics. But if I only went by stats I would have probably went with a 24-70mm or 24-105mm because most of my shots were not at the 24mm end at all. But I knew that was when it was snowy and I'd likely get more sweeping vista's. And there was some some shots I ended up doing a pano, but I still wanted more width.

I eventually opted for an ultra wide, which went against what my stats were saying. Because sometimes stats are not everything with lens selection. I knew well enough what I should be getting, rather than just stats.

I more or less dropped the 24mm f/1.4 not only because of focal length, but just as much because it was the original version available locally and I wanted the Mark II.

And after going there and using the 16-35mm it was the correct focal length range, if anything I could have went wider on some shots. I used it most of the time the 16mm end, but I did get some of my favorite shots on the 35mm end as well. So the focal length range proved to be very useful to me. And it didn't come to a surprise because I had the Tamron 10-24mm on the 60D I used to have and enjoyed immensely that lens and focal length range. It really was a great lens though if anyone ever wanted something like that, I was pleased with the IQ of my copy and obviously keeping in mind price and intended format.

I did the same thing with the 24-105mm I had (I owned both versions.) vs the 24-70mm. Statistically I like you used both ends of the focal length range quite a lot. I used IS quite a lot. But I still wanted a 24-70mm. So I got a Sigma version, I was in love with the focal length and aperture combination! So much better for me. Do I miss the longer end? Yeah, I do. But having f/2.8 was more important to me and transformed it for me. So much happier with that combo. But I sold that and now want the Canon Mark II.

Needless to say, sometimes you have to make a decision contrary what your past use experience because while you might have made use of the previous focal lengths, IS or aperture. You will shoot differently with the 24-70mm compared to the 24-105mm. And only with using them side by side will you know which is the right one for you.

Quote:

---

Originally Posted by neuroanatomist

If I had to bet, I'll use the 24-70 more but take the 24-105 when I travel.

---

That makes a lot of sense to me. In all the analysis, my primary thought beyond where it is used, is what could you live without? The missing 35 mm at the long end of the 24-70 is easy to compensate for by cropping. Sure, lost of pixels, and noise becomes more apparent if at high ISO, but it can be done with minimal penalty. Plus, I often have another lens in the bag that could be used to compensate.

But f/2.8 vs f/4. If you need that stop of light, if you want that bokeh, there is not an easy replacement.

Ah man the GAS is itching bad.

I have loaded up the shopping cart a couple of times. Like the idea of smaller(ish) cards vs the giant capacity. I had to giggle at the size of my hard drives vs the size of the cards. My laptop is 500g ssd an upgrade a few yesars ago., the desk unit is 1T. Space is reasonably cheap really tells me how long I have been happy with the core processors vs storage capacity of cards. Just an odd juxtaposition you might find humorous

I thought this was interesting...I do not know these cards...but they are no publishing a "minimum write speed" that is pretty consistent with what test results people are finding.

Attachment 2930

Wish my CFexpress cards are getting anywhere close to those numbers transferring to my PC.

I know what the issue is, my computer doesn't support USB 3.2, so I'm just limited to 37/40MB/s.

I have been wanting an upgrade for a while now, blemish removal is incredibly slow in DPP. Everything else is great. I have to wait forever per image to load and then proceed with my work. And even then it is extremely slow. Regardless of the resolution, even if it is just my 10mp 1D III images it takes FOREVER!

Yeah...generally this has been my surprising conclusion as well, that CFE cards are well below their listed performance. So, all of the theoretical calcs that those of us do, myself included, using the list specs are somewhat erroneous.

Hopped back to this thread as Camnostic has updated their review, for those interested. I still prefer what SSDreview is doing, but Camnostic reviewed specifically ont he R5, so some relevant data there.

Overall, having now shot on the R5 for ~18 months, what I can say is that I cannot recall ever hitting my buffer when writing to the CFE card (I have Sandisk 128 GB/512GB cards). But I have hit it when writing to the UHS II card slot. My solution has been to write only to the CFE card when shooting anything at high FPS and then copying over to the UHS card when I have a down moment. For lower FPS action, I write to both cards simultaneously.

So, while CFE cards do not perform at their spec, even when transferring files, they are still more than fast enough for my needs. The Sandisk CFE cards do get hot. If I was to buy another card, the operating temperature chart would get some of my attention, in addition to speed. This may seem odd, but after transferring files, the Sandisk cards are hot enough, I let them sit before reinserting into the R5. They can also be very warm when pulling from the R5.

I ran into an issue where the original CFExpress card I purchased -- a ProGrade Digital 128GB CFexpress 2.0 Type B Gold -- wouldn't allow me to record 1080p 120fps, let alone 8K RAW video (it would stop recording after a few seconds). I had originally planned to get a Delkin Devices 128GB BLACK CFexpress Type B but found an ebay deal on a used Delkin Devices 128GB POWER CFexpress Type B (I paid about $100) and thought I'd give it a shot. It's allowed me to record high framerate 1080p video and 8K RAW without a hiccup and I've been very happy with it. If I hadn't found the deal on the Gold (a lesser-tier) card, I'm sure I would have been happy with the Delkin Black series card which, at this time, has 46 5-star reviews (no reviewer has rated it for less than 5 stars).

@Kayaker

I use two SanDisk 64gig cards, I never noticed them getting especially hot in practice. Do they get warm when under heavy use like when using them on my 1Dx III? Possibly, but I never pull them out right after heavy shooting. I pull them out when I get home to load the images.

And of course I'm limited in my upload speeds because I don't have USB 3 2.

Not to say I they don't get warm, just my personal experience using them.

Quote:

---

Originally Posted by Fast Glass

Do they get warm when under heavy use like when using them on my 1Dx III?

---

Of course, I am not sure what your heavy use would be and heat disappation is likely different between the R5 and 1DX III. I, admittedly, have a heavy trigger finger. I have removed my CFE card from the R5 after heavy use and I'd describe it as warm.

I would describe it as hot after it had been in my card reader for awhile and transferring files. Somewhat interesting, just being in the card reader makes it warm to hot (the computer recognizes it as an SSD drive). Transferring a couple hundred GB of files, it is definitely a hot. In the figure I referenced above, theSSDreview found that the 512 GB Sandisk card got up to 80C during large file transfers. I wouldn't say mine was 80C, but the trend is there. It has interested me that a number of cards are much cooler under the same conditions.

The 64 GB Sandisk card has different list read/write speeds, so I am not sure if it would be directly comparable to the other Sandisk cards. But, in terms of heat generated, maybe that is a good thing.

Quote:

---

Originally Posted by Kayaker72

Overall, having now shot on the R5 for ~18 months, what I can say is that I cannot recall ever hitting my buffer when writing to the CFE card

---

Just to clarify, you're referring to shooting in mechanical shutter mode, correct? I find that the buffer does max out in electronic shutter mode, but that's not a surprise at 20 fps.

Quote:

---

Originally Posted by Jonathan Huyer

Just to clarify, you're referring to shooting in mechanical shutter mode, correct? I find that the buffer does max out in electronic shutter mode, but that's not a surprise at 20 fps.

---

I do usually shoot mechanical. So my experience is skewed towards mechanical. I do occasionally shoot electronic, but when I do I am trying to time specific moments and shooting in bursts.

The camnostic article had 124 frames before hitting a stutter with the R5 and Sandisk 512 GB card at 12fps (mechanical H+). Bryan had similar, but slightly higher, results on a different card. So it is possible, I just haven’t had a lynx walking down a road toward me yet [emoji6].

I am pretty sure you use Sandisk 512 cards as well. Camnostic’s test had ~310 images in 30 seconds in e-shutter mode. The best cards were around 340. So not much of an improvement to the fastest cards. But then the Sandisk 128, which is my other card, was “only” around 260 in 30 secs. So there can be a difference if you need this type of performance.


Sent from my iPhone using Tapatalk

Ah that's interesting. I also shoot the R5 in mechanical mode, and yes I have Sandisk 512 cards. I checked one sequence I did with a lynx walking towards me, when I leaned on the shutter and didn't let go until the cat was gone. It lasted 43 seconds and I got 207 shots. There was one quick 2 second interruption, when I adjusted my exposure compensation (love that you can see when you need to do that!). So overall I got 5 fps. I was shooting at f/5.6 with the 500 mm lens and 1.4 TC attached, so the aperture was wide open. But maybe the fact the cat was moving towards me slowed down the shot rate a bit, as the camera has to adjust focus every time (?). I am pretty sure my battery was close to fully charged, since the encounter happened in the morning without any other prior shooting. This 'real life test' shows that there's room for a faster frame rate, since I never came close to the ideal-world 12 fps. Not that I'm complaining --- I got a pile of keepers from that encounter :cool:

https://photos.smugmug.com/Nature/Wi...987-Edit-L.jpg

Amazing shot.

Looking at the Camnostic data, it does stand out that when holding down for 30 seconds, the R5 performance for most cards was fairly identical whether they were in e-shutter or mechanical, indicating that it is really the write speed to the card that is limiting for longer bursts. I'd put the Sandisk 512 right around the same 310-315 both in e-shutter and mechanical.

Borrowed from Camnostic's data:

https://camnostic.com/wp-content/upl...n-eshutter.png

https://camnostic.com/wp-content/upl...urtain-all.png


Overall, I would say 207 images in 43 seconds is a bit off. I wonder if your battery, even if fresh, was cold? Maybe card speed declines as you have more and more photos on it?

I have focused on 3-5 second bursts when I tested my R5. I hit 20 fps, 12 fps, 9-9.4 FPS pretty consistent in those short burst for E-shutter, H+, and H modes and full battery. One complaint I have with the R5 is that FPS in mechanical shutter modes does decline, noticeably so, as the battery levels drop. I am hoping this is fixed in the eventual R5 II, if not, I do consider this a significant advantage of the R3.

Yes something was probably off. The battery is the best guess, although it was warm inside the vehicle. I just did a test at home with the same setup, using a fully charged battery. In the first trial, I got 238 images at 9.2 fps before the buffer filled. I moved the camera around to simulate a moving subject. Next I switched to manual focus to see if that would make any difference. I got 170 frames at 9.4 fps average, so not much of an impact. The ISO setting will definitely have an impact on things, because higher ISO results in larger files. In my test, I was at ISO 3200 and my file sizes were about 59 MB. For the lynx shots, I was at ISO 800 and the file size was about 50 MB.

Next I switched to C-RAW, and the file size dropped to about 30 MB. I could maintain 9.2 fps indefinitely without filling the buffer (I gave up after >400 shots). Finally I tried electronic shutter with C-RAW, and I got 98 images at 20 fps.

I can't seem to get more than 9 fps with a mechanical shutter, no matter what I try. I followed all the guidelines in the manual (full battery, compatible lens wide open, 1/1000 sec shutter, no flicker detect) and it won't go any faster.

I also noticed that the drive mode icon remains white, indicating 9.2 fps. So at least it's telling me the truth. However I can't make it go green, which would indicate 12 fps is enabled.

Meanwhile, I think I will stick with C-RAW. Everything I've been seeing about it says it has no impact on image quality. And if it gives you unlimited shooting and smaller files, why not?

I have elected to shoot electronic shutter, 68B colors (12 bit) is deep enough for me.

I haven't leaned on the shutter for more than 3 or 4 seconds at a time. I have an anglebird 512g, AV Pro. Haven't noticed anything other than 20 fps so far.

Is the H+ green when you shoot it? If it is white, then some setting is preventing you from 12 fps.

Attachment 2990

I just did a couple tests, with it white, I was at 9.1 fps twice, once with a battery at ~12% and again with a battery that was 97%. But the Icon (H+) was white. So, going through the list, my issue was my wifi, while not "on" as in connected, I did not have my camera in airplane mode. I had connected my camera to my phone via bluetooth recently and that was enough. As soon as I turned my camera to airplane mode, my H+ was green, and I just ran another ~5 second burst, and averaged 12.0 fps over that burst.

Airplane mode is wifi tab #1.

Ah ha! That did it! Thank you!! I just turned wifi off and now I have a green icon. I did a test and got 247 shots in 20 seconds for 12.3 fps before the buffer filled (in C-RAW).

So if I could recreate my lynx encounter once again, I think the best setting would have been 9 fps in C-RAW. Then I could shoot constantly for the full 43 seconds (and beyond) without hitting the buffer. For sure, 12 fps for 20 seconds is pretty nice, but I've had several experiences besides this lynx where it would have been inadequate. For example, bear cubs play fighting (they'll do that for a couple minutes easily). The 12 fps mode is a lot better suited for birds in flight, where you would never need 20 seconds worth of shooting. The next time I go out for owls, that's what I'll use.

Dreaming ahead to the mythical R1, I think it would be good if it comes with a lower MP count than the R5 (say, 30 MP). Then it would stand a better chance of being able to write at 12 fps without buffer issues, especially in C-RAW mode.

More testing: With C-RAW, 1.6 crop factor, and electronic shutter, I got 306 shots in 15 seconds before the buffer filled. I kept the button down and the shutter paused for 3 seconds, then resumed shooting for 4 seconds, in a repetitive cycle. After 30 seconds, I had 481 images.

I know that most people really don't shoot this way, but there are definitely times that I do, and those times are always the most exciting moments in wildlife photography. So this has been a good exercise in finding out what the best mode would be that can allow continuous unlimited shooting. The answer for me is 9 fps in C-RAW, and that will now be my default unless I am going birding.

This really does not seem to pertain to me, but it might someday, if I am shooting a lynx as it walks towards me. But this past month I was shooting birds with Mike down in Florida and just got back from vacation in Costa Rica where I had two half days primarily aimed at photography. With the birds, I was using 20 fps. But, even most of the time in FL and all the time in Costa Rica, I was shooting in "H" mode, namely, ~8 fps. I never hit my buffer in either instance.

That said, I started to wonder if it has been long enough that CFExpress Type B cards are starting to match their specified write speeds as the Sandisk ExtremePro 128 GB and 512 GB cards I have do not....far from it really.

So, for those of you with a need for speed....it looks like the latest cards are now 2-3x faster than those I bought in 2020.

Attachment 3029


Then, they are also running cooler.

Attachment 3030

The article indicated that the Lexar Pro Diamond 512 MB, while best performing, they expect the price to come down from the $530 point it is now. Admittedly, the Prograde Gold is right in there in terms of performance and is $179 for 512 GB and $400 for 1 TB. Previously, I have stayed with Lexar or Sandisk, but Bryan is using Prograde in his reviews. Right now, I'd likely go that direction.