Shirozina

I've been using an eGPU for a few weeks now on my XPS 9560 and love it despite the extra bulk but you do have the option of doing basic stuff without it and then plugging it in when you want to get down to business. The current generations of mobile CPU's and GPU's are good enough for 4k editing in terms of processor speeds but when put in thin laptops they generate too much heat to be used for anything other than short busts of use.

Yes but I think the test was flawed and my reasons are in the other thread.

I don't see how the Mac is a lot better unless you want to use FCP as they are basically sharing the same hardware ( XPS slightly better in the GPU dept) and both limited by thermal design to doing non intensive tasks.

The Dell does throttle the CPU on battery but I have to ask who is actually going to be doing any serious editing just on battery? There are various battery settings in both windows and the Dell power command app that I have no idea if he set up optimally. Also there are various settings in the Nvidia contol panel which effect how fast it works with video processing and no indication if these were optimised. The basic problem with these laptops not addressed here is that they will both throttle when set to do longer processing tasks like rendering cache files or optimised media which are common tasks you need to do when working with compromised systems and having to do workarounds (Laptops being generaly inferior in CPU and GPU etc to desktops) and throttling is totally unavoidable in laptops due to the physical impossibility of cooling in a thin design. Whatever you choose - Mac or PC I'd say avoid the top end i9 models and spend the money saved on an eGPU system which will mean you can (with Resolve especially) do lengthy high intensity processing tasks without throttling issues.

Yes ( Sorry I should have mentioned that). These boxes take normal desktop size GPU's and do vary in size as some have internal PSU's. The Asus XG station pro is not too big and has an external PSU. It's also not that expensive and put in a GTX 1070 or above and resolve will fly.

If you have the space get an eGPU as this will make virtually any reasonabe spec laptop run Resolve OK.

I just enabled the ignore user feature for him but be aware that not everyone is 'neuro-typical' and that if you are on-the-spectrum this can come across as plain rude and aggressive on a forum.

I didn't make any claims about the GH5s - again you really need to read my posts more carefully....... You are talking about 1 camera. I referred to 'large sensors' as a general concept which hold true for the vast majority of cameras without Dual ISO capability.

Even Better DR is not a given as the GH5s shows with it's ability to maintain a good DR even at high ISO's. Large sensors may have less noise s you crank up the ISO but the DR drops off quite a bit with it.

Not used the GH5s but shooting video on the GH5 at base ISO I don't see noise unless I underexpose or have to boost the shadows. Even then the quality of current noise reduction software like NeatVideo and Resolve Studio really don't make noise a factor I worry about. BTW - maybe you could post a screen grab to show this noise visibility at all ISO's? - or maybe not

But not every situation demands you push the cameras ISO to it's limits - within base or near base ISO there are no practical differences in noise that would make you choose one snesor size over the other

You are comparing sensor tech not sensor size. Once you get down to base ISO levels where noise is not present there is no advantage other than DOF differences and as the Dual ISO in the GH5s proves sensor size doesn't necessarily mean poor higher ISO performance. BTW - given the sensor size difference between the GH5 and the A7r3 the visible difference is not as large as many would assume....

In the world of 'videography' as opposed to actual commercial film making in order for you shots to be concidered as 'cinematic' you have to have shallow DOF no matter what......

Esp when they don't understand the concept of ISO as it applies to LOG capture profiles.

Very true.

The problem with fast lenses on M43 that achieve similar DOF characteristics to S35 and FF is the amount of ND filtering you need in anything other than low light conditions. If you are a Vari ND user you will easily run out of density and need to add a fixed ND on top. You are then stacking 3 layers of glass so they better be high quality with good multi coatings so that 6 air to glass surfaces are not going to degrade your image. Internal ND filtering would be nice but not sure how this could be achieved in a small body like the GH5s where they already have had to ditch the stabilizer.

By definition a 'screen grab' is likley to be a poor method of demonstrating this even if I had side by side shots of the same subject which I don't. I also don't need to do such comparisons to get a general impression of how good a camera is as this comes from working with multiple clips of different subjects over time.

The screen grab tells me nothing other than Sony cameras have OK image quality which no one is disputing. I didn't say 10bit was the only factor - have you read my post? What 'money' do you need to see exactly?

The GH5 internal codec is technically superior to the the Sony one as it has more bit depth, more chroma subsampling and a higher bitrate which gives the end user substantially more room for tonal and colour changes in post production and makes capture with a LOG or HLG gamma profile a practical proposition with non of the problems this has with the Sony internal one. I don't care about your sample image (which demonstrates what exactly?) and am only concerned how my footage looks and having shot quite a lot with both codecs I know what I prefer. The Sony sensor has advantages ( DR and high ISO for eg) and in stills ( RAW) the Sony is better but it's held back in video by it's poor internal codec which is also evidenced by how much better it behaves when you use an external recorder and can improve the bitrate and chroma subsampling.

I'm talking about the codec and not the DR, noise characteristics or res of the sensor and lens - too many variables.

I find the M43 sensor size too limiting for general video use in that it severely restricts DOF options without adding exotic ultra fast glass ( primes only mainly) which then have to be used with heavy ND filtering in anything but low light. I do however think that Panasonic's 10 bit 4.2.2 codec is significantly better than Sony's 8bit 4.2.0 - it just looks better with more subtle tonal separation and can be used with LOG picture profiles without worry it will fall apart in grading. If Sony match their internal codec to the same for the A7sIII I'll probably jump right back. On the whole I've not looked back since I ditched my 5DII for an A7r - mirrorless IS the future and DSLR's are on borrowed time.

The whole concept of ISO and 'exposure' don't apply to LOG - they are leftovers from stills profiles. ISO for instance is measured at a sensitivity in the middle of the response curve which in a still or non log picture profile is flat but in a LOG profile it's highly elevated so '1600' ISO is meaningless as it describes a point on a highly elevated curve. Same with exposure meters which are designed to zero on mid grey on a stills type response curve when on a LOG curve it's again highly elevated. To expose LOG properly you need scopes.

So basically a near ETTR exposure - horrible banding / colour clipping around the light clouds towards the top right of the frame. the 8bit 4.2.0 100mps internal codec is just not up to this kind of low contrast scene. You would probably get away with it shooting to an ext recorder in 4.2.2 with a less compressed codec but to capture a low contrast scene like this internally you need to pick a more contrasty profile to spread your tones across more of the image data range.

I think you are only interested in arguing but here you go; 1, You can't invent colours that are not captured or recorded in the first place. I get paid to manage colour but I'm not a magician 2, REC709 and many other non log camera profiles have limited DR and Gamut so it doesn't matter how many bits the file has if the data isn't there. 3,If you view rec709 on a REC 2020 device without transforming it that isn't a workflow it's called operator error. SD to HD - what point are you making in relation to any of the above?

Who is going to transform REC709 to REC2020 as a workflow? Most will be doing the opposite so with the correct transformation no data will be lost. Log is not a problem as long as it's captured using enough bits and enough chroma subsampling. The problem is when you shoot LOG in 8bits and 4.2.0 and then on top of that use high lossy compression codecs.