Canon 5D Mark IV -- Time to Discuss!

CyberDyneSystems wrote in post #18215901
Ditto

Maybe we have different ideas of slow?

Likewise, ... still slow,

4 Core 4Ghz Intel, 32GB RAM, and SSD, still slow! (compared to other RAW converters I use)
Back when I had my AMD Opteron set up, LR was still the slowest, and RawShooter Professional (R.I.P.) was lightning fast by comparison. Likewise Bibble (also R.I.P.) and C1.

I don't think it is strange, it seems to be the norm based on a combination of different versions, (going back to 1) different installs (at least half a dozen by now) and all the similar reports I have read over the years.

My favorite part was when they added GPU acceleration and turning it on made it slower

Jake - As for different ideas of slow I hear a lot of complaints about lag but I don't see it. When I move a slider, use a brush, use a preset, change modules, etc. my results are as instant as they have always been. What are you experiencing?

LR mostly runs fine for me - 4790K, 32 gb RAM, OS, programs and catalog on SSD, Win 10 Pro, GPU enabled.

However, every now and then when I used alot of adjustment brushes and / or spot removals, LR slows down noticeably. If I exit the program and reboot, it runs fine again. So I think there's a memory that gets reserved but doesn't get released after use and it builds up slowing down LR. The reboot clears the memory.

Mine runs very well on Win 7 SSD 6 Core 5820K but there is a memory leak in it. So I restart and don't leave it running when I'm not using it.

jonneymendoza wrote in post #18215098
LIGHTROOM IS SOO SLOW!!!!!!!!!!!

I scoff at you slow complaint

I am using an i5 with 8Gb of ram because I am too cheap to buy a better PC.

If I let it do it's thing, import, build previews and complete (go for a 5 K run while it's doing that) it's usable enough for me. I store the the images and the catalog on a removable hard drive.

I would much rather spend that money on glass.

...yeah, well...and it still slows down for some things, particularly for 5DsR files...

- Intel i7 6700K CPU @ 4Ghz - Noctua NH-U12S HSF - 64GB Corsair Vengeance 3000 RAM - Samsung 950 M.2 NVME 512GB boot drive - Gigabyte Z170X 7 EU MB - Nvidia Quatro K2200 GPU - Samsung 850 Pro SSD 512GB - Samsung 850 Pro SSD 2TB - Samsung 850 Pro SSD 1TB - Antec 850W PSU - SilverStone TJ07 case - 8TB Synology NAS

I agree on the not releasing memory, plus not using the full extent of available hardware. Either it is some obscure marketing strategy, holding back on development costs, or both.

LJ3Jim wrote in post #18210933
When I look at his 100% crops comparing the 5D4 to the 1DX2, I understand that the 5D4 puts "more pixels on target." Thus the crops from the 5D4 enlarge the birds 22% compared to the 1DX2. However, in both comparisons (the hawk and the cormorant), I think the 1DX2 shot looks better. It's almost as though the quality of each 1DX2 pixel is slightly better than each 5D4 pixel. Any agreements or disagreements?

To my eyes, I have always preferred the way larger pixels look at 100% crop level. Saw this when I rented (before buying) the 5D3. High pixel density sensors (again to my eyes) look really good when uncropped but I dislike the per pixel look they offer. I realize each of us sees images differently and my statement here is valid only for my eyes. I also try not to crop to 1:1 levels when I can avoid it.

I currently have a 5D3 with (hopefully) a 1DX-2 coming for Christmas. At some distant point I may replace my 3 with a 4. Looks promising early on!Just saying this as I have no bias

jdavis37 wrote in post #18216109
To my eyes, I have always preferred the way larger pixels look at 100% crop level. Saw this when I rented (before buying) the 5D3. High pixel density sensors (again to my eyes) look really good when uncropped but I dislike the per pixel look they offer. I realize each of us sees images differently and my statement here is valid only for my eyes. I also try not to crop to 1:1 levels when I can avoid it.

I currently have a 5D3 with (hopefully) a 1DX-2 coming for Christmas. At some distant point I may replace my 3 with a 4. Looks promising early on!Just saying this as I have no bias

At the risk of getting off topic here, I've seen a bit of both.

I've seen the same e.g. G10 vs 40D and A7R vs D4

However, comparing D7000 (16mp DX sensor) to the D700 (12mp FX sensor) I replaced it with I didn't see the improvement I was expecting from the larger sensor. This case may have been affected by the different generations of sensors however - that D7000 Exmor was a very nice sensor even if it was only a DX size.

jdavis37 wrote in post #18216109
To my eyes, I have always preferred the way larger pixels look at 100% crop level. Saw this when I rented (before buying) the 5D3. High pixel density sensors (again to my eyes) look really good when uncropped but I dislike the per pixel look they offer. I realize each of us sees images differently and my statement here is valid only for my eyes. I also try not to crop to 1:1 levels when I can avoid it.

The problem of course is that you're not comparing like for like; if you took a 20MP sensor, used a good algorithm to downsample it to 5MP (basically each 2x2 grid of pixels becomes 1) then it'd likely look better at 100% than the original (partly because you've averaged and smoothed the noise).

However, if you printed the 20MP image and the 5MP image side by side, at a size where you couldn't see the individual pixels in the 5MP image, then at worst they'd look the same; more likely the 20MP image would have more detail.

The higher resolution image would also allow a larger print size (or more cropping) before individual pixels became visible. There would be a limit of acceptable noise in the print, but that's more about noise per unit area than resolution. I.e. a lower res sensor with the same noise per unit area would have a similar limit - if not an earlier limit; due to the individual pixels becoming apparent.

One other way to look at it - a 100% crop of a higher res sensor isn't the same as a 100% crop of a lower res sensor; it's a higher zoom/tighter crop (which would likely be unfavourable to the low res image).

sploo wrote in post #18216997
The problem of course is that you're not comparing like for like; if you took a 20MP sensor, used a good algorithm to downsample it to 5MP (basically each 2x2 grid of pixels becomes 1) then it'd likely look better at 100% than the original (partly because you've averaged and smoothed the noise).

However, if you printed the 20MP image and the 5MP image side by side, at a size where you couldn't see the individual pixels in the 5MP image, then at worst they'd look the same; more likely the 20MP image would have more detail.

That is all fine. Until you start taking advantage of all that nice resolution and cropping the image hard, e.g. deciding not to use a longer focal length (not available, too costly or just more convenient not to change lenses) because you have lots more resolution to give you digital zoom.

You might complain that this is an apples vs oranges comparison but it is a not uncommon occurrence - there have already been a number of discussions in this thread alone around the relative merits of using an 80D or 7D2 for better digital zoom vs using a TC on a 5D4 for optical zoom or sticking with a shorter lens on an 80D or 7D2 vs shelling out for a longer lens on the 5D4.

smythie wrote in post #18217040
That is all fine. Until you start taking advantage of all that nice resolution and cropping the image hard, e.g. deciding not to use a longer focal length (not available, too costly or just more convenient not to change lenses) because you have lots more resolution to give you digital zoom.

You might complain that this is an apples vs oranges comparison but it is a not uncommon occurrence - there have already been a number of discussions in this thread alone around the relative merits of using an 80D or 7D2 for better digital zoom vs using a TC on a 5D4 for optical zoom or sticking with a shorter lens on an 80D or 7D2 vs shelling out for a longer lens on the 5D4.

An then using that longer lens on the 7D2


Now that I have the 5D IV , I don't even touch the 7D (classic) anymore :/

It is going to warm up so hopefully I'll get out and find some birds and check out my 100-400 II and 1.4 III.

smythie wrote in post #18217040
That is all fine. Until you start taking advantage of all that nice resolution and cropping the image hard, e.g. deciding not to use a longer focal length (not available, too costly or just more convenient not to change lenses) because you have lots more resolution to give you digital zoom.

You might complain that this is an apples vs oranges comparison but it is a not uncommon occurrence - there have already been a number of discussions in this thread alone around the relative merits of using an 80D or 7D2 for better digital zoom vs using a TC on a 5D4 for optical zoom or sticking with a shorter lens on an 80D or 7D2 vs shelling out for a longer lens on the 5D4.

It's likely you should get a bit of advantage in cropping/zooming with the higher res sensor - but sure, as a replacement for a longer focal length (or a TC) then it's not going to work that well. That said, taking a 1.6x crop from a 5Ds is (in terms of sensor area, glass coverage, and resolution) almost exactly like using the same lens on an APS-C body.

Point being: higher res is almost certainly never going to give you a worse result (same framing/same print size/same glass, vs a low res sensor), and has a decent chance of giving you something better.

sploo wrote in post #18217346
It's likely you should get a bit of advantage in cropping/zooming with the higher res sensor - but sure, as a replacement for a longer focal length (or a TC) then it's not going to work that well.

Smaller pixels can easily be superior to a TC. It really depends on the specific cameras, as we are talking about noise per unit of sensor area, which does not vary by a huge amount in many of the cameras of the last few years. The 6D, 5D4, 1Dx, and 1DxII are the only Canons that have less noise per unit of sensor area than the 7D2 (the 6D and 5D4 by the least amount). Those cameras, with a TC, outperform the 7D2 without a TC, in terms of subject-level noise (the 6D and 1Dx are inferior at low ISOs, though). The 7D2 without a TC has less extra glass to put slight halos at edges or reduce global contrast a hair. Of course, it depends on the specific model and AF points needed how autofocus will compare with a TC against a bare 7D2. BIFing fares better with the all-point f/8 cameras, not so much center-point-ing, where the limit of the 7D2 allows more pixels-on-subject with the same TC or number of TCs.

Don't forget, shooting at the same shutter speed with a 1.4x requires a doubling of ISO! You can't use the fact that FF sensors have less image-level noise at the same ISO in general, unless you drop the shutter speed to keep the ISO down. As far as pixels-on-subject is concerned, only the 5D4+1.4x beats the bare 7D2, but only by putting 1.18x as many pixel on target.

Question for 5D4 owners.... when I bought my 5D3 I will say upfront overall it was the best camera I had owned and while it had some compromises compared to the 1 Series, in general it was a perfect fit for me budget wise, performance wise and so on.

Going forward the 5D3 will be a second body for me (backup for failure, second body to avoid lens changes between telephoto and wide angle, etc. ). This is assuming Santa delivers a 1DX-II on Sunday.

My question though is in regards to the 5D4's frame rate. I have read the true buffer size has been significantly increased over the 5D3's (which is a good thing). Frame rate.. my 5D3 has never felt like a "fast 6 fps" body. At full battery charge it approaches 6 fps but not sure it quite gets there. By time batteries have discharged to 60%, the frame rate has dropped to about 4 fps. When shooting birds the slower frame rate can become a problem (for landscapes going forward not so much of one obviously). Does the 5D4 also drop frame rate significantly as the batteries discharge or do the newer batters, etc. help prevent that?

I am considering at some future point selling my 5D3 and replacing it with a 5D4 (year or so from now) to take advantage of the lower ISO DR and higher pixel density for my landscape work. DX2 will be my birding camera of choice, action work with dogs, and so on. But I can almost make a justification for going to the Mk4 and replacing the 3. If the frame rate and buffer are both significantly improved it means the backup role will also be improved. Thx

I have not noticed a drop in frame rate, but that's just impression based on use/memory. I must admit I did not notice the drop in the 5D3 so my impression is obviously of little worth.

On paper, the only thing holding the 5D4 back is the small size of the buffer,. (relative to a 1D) the flush to card is pretty peppy. That would give me reason to assume that it does not slow down... but again, not an actual test.