IMPORTANT UPATE REGARDING NOISE AND SENSITIVITY: See section highlighted in red below.
Hi all. Well I have got a Canon XF305E for the afternoon. Wish it was longer, but they are like hens teeth. I’m going to be writing and updating this as I go, so please keep coming back for the latest updates and post a comment if there is anything you want me to specifically look at.
Out of the box, first impressions are that it is big. 20% bigger than an EX1. The body is dominated by the very large lens which also has a sensor of some kind to the left of the lens barrel. I assume this is to do with the autofocus system. I hope so as it would make Matte Box use very difficult.
Overall it feels very well made although there are a few bits that could be better. It’s possible to put a battery in the battery compartment incorrectly so that the camera will work unless you knock it and bump it, then the battery connection is lost. You really need to have the camera down on a flat surface to be sure the battery goes in right. In addition there is a pair of really heath robinson looking springs at the back of the battery compartment. Considering this is about the most expensive camcorder in it’s class you don’t expect cheap and nasty springs like the ones fitted in the battery compartment.
They look like an after thought.
Incorrectly inserted Battery
The flap that covers the battery feels cheap and plasticky compared to the rest of the camera body and I would be worried about this breaking off at some point. However if it does break there is a separate battery catch that holds the battery in place.
Looking at the lens my first point of confusion came when I pressed the zoom rocker and nothing happened. I turned the calibrated zoom ring on the lens and the lens zoomed in and out, but the rocker did nothing. So I looked for the zoom servo switch. I didn’t find one but did find the “Zoom – Rocker – Ring” switch. Switched it from ring to rocker, pressed the rocker and the zoom works…….. but….. now the zoom ring does not turn or move, so you can have one or the other but not both. The iris ring has no markings and is of the round and round servo variety so you need to look at the (very nice) LCD screen to see where the iris is set, the same for the zoom when set to rocker.
The focus system is not that dissimilar to an EX with both calibrated manual and non calibrated manual/auto operation. On an EX you slide the focus ring forwards and backwards to switch from full manual to servo control. On the XF305 you have to take your hand off the focus ring to operate a small push button that allows you to rotate the lens body to switch between manual and servo. It’s not quite as convenient as the EX but is easy enough to do. There is a problem with this however. If you set the manual focus ring at say infinity, then switch to the servo/manual/auto focus and use that, when you switch back to full manual the lens will return to the last position you set the focus to. OK, fine, that’s the same as an EX1R, BUT on the XF305 you can’t see the focus scale in the focus window when you are set to servo/auto so you have no idea what the focus is going to do when you switch modes unless you can remember where you last set the manual focus.
Just like an EX1 when you hand hold the camera it’s bulk puts a fair bit of strain on the wrist as it want’s to fall forwards and to the left. There is no adjustable hand grip rotation and the record start stop button is a little low down for my thumb. The hand grip is also quite small and angular. I think the EX1R is a lot more comfortable to hold.
If your using the XF305 on a tripod there is only a single 1/4? threaded hole in a small plate on the bottom of the camera, much like the original EX1, so don’t overload it or you may end up breaking the plate. Perhaps Curtis at Juiced Designs will do a strengthening plate for the XF305. Do camera designers not read forums and look at what happens to cameras when used in the field?
Behind the hand grip there is an array of BNC connectors for SDi, video out, Genlock and timcode. Full marks for using proper BNC connectors here. Behind these are a whole bunch of floppy cheap plastic covers over HDMI, USB, headphone, AV, remote and mini component connectors.
In use the large LCD screen is clear and easy to see. It flips out from under the handle on both sides of the camera which is really neat.
The multi-coloured graphics all over the screen do make it a little cluttered but these can be easily turned off. The XF305 can also display waveform monitor and vectorscope plus a 3 zone waveform monitor all of which are very useful tools to have (LCD screen only, not rear VF). While looking at the LCD I realised that just like the EX1 the microphone and in this case the LCD as well, stick out beyond the end of the lens. Why do camera manufacturers do this? It makes fitting and using a Matte Box so difficult. Doh! Having the LCD screen so far forwards could present problems for Matte Box users.
Where the EX1 and EX3 have picture profiles the XF305 like most pro Canon cameras has a number of “Custom Picture” memories. There are 6 preset memories for you to dial in your own looks plus “Video C”, “Cine V” and “Cine F” setups. The Cine V setup was really soft and Cine F just had a quite flat look, however I didnt really have time to explore these fully. I did try Alan Roberts recommended settings and these are very nice and I would recommend them as a good starting point for really making the XF305 sing. Within the CP settings you can choose from 4 standard gamma curves and 2 cine style gamma curves. The Matrix is fully adjustable so it should be easy to roll your own custom looks. Another setting tucked away in the CP menu is the noise reduction. The XF305 has some very clever noise reduction that is clearly doing a good job of controlling the noise that is normally be associated with a small sensor camera. You can choose between Automatic nose reduction and 8 steps of reduction. If set to 8 however the noise reduction is very hard and the resolution drops way down.
The pictures from the XF305 are very good. At 0db they are quite similar overall to those from an EX1. They do have a different colorimetry to Sony’s camera’s (which I always find a little yellow) and are pleasing to look at. Clearly the lens is very good, CA is well controlled but there is some quite obvious barrel distortion between fully wide and 6mm, but it’s no worse than an EX1R and really to be expected from a camera at this price point.
I also found some noticeable softening in the corners fully wide which appears to be worse on the left side of the image than the right.
When I was looking at the lens distortion I was zooming in and out using the zoom rocker, then I went to turn the zoom ring just to tweak the zoom and of course nothing happened. This zoom rocker or zoom ring, but not both way of operation really sucks.. am I missing something here? While on the subject of the zoom rocker I was asked about how it was with slow creeping zooms. Well it’s very good. I did find that you have to press the rocker a long way past the center point before anything happens, but once the zoom starts to operate you can get a slow creeping zoom. But to then go from the creeping zoom smoothly to a faster zoom is tricky as a tiny bit more pressure on the rocker leads to a rapid increase in zoom speed. It seems that there is a very large dead area where the zoom rocker does nothing and then all the action takes place in the last few millimeters of its travel. It almost feels like the zoom speed is in steps, not entirely variable. I’m sure it’s not and with practice perhaps I could master it, but it’s a little touchy. The EX1R isn’t perfect either. The zoom rocker can be a bit twitchy when trying to do a creeping zoom, but in this case I prefer the EX1R zoom over the Canon. If you do choose to use the zoom ring to control the zoom there is some serious lag between turning the zoom ring and the zoom happening. For slow zooms this may not be an issue, but crash zooms are very difficult to execute. The iris ring also as some lag which makes fine tweaking of exposure a little harder than it should be. The last couple of years I have become so used to the EX1 and EX3 lens with its great feel and proper zoom and iris rings that this is a real let down. Overall I think that optically the Canon lens as a small edge over the EX lens, there is less CA and a greater zoom range. But ergonomically I much prefer the EX1R lens.
Latitude is very similar at 0db between both cameras, I couldn’t really see much of a difference either way. UPDATE There are some differences in image sharpness and noise however. Please see this clip (Vimeo or YouTube) for some examples, generally Vimeo is better quality.
If anything the Canon 305 at first glance appears a little sharper than the EX1R, but looking closer and examining the footage as well as resolution chart results shows very similar resolution from both cameras. However the XF305 pictures contain a lot of quite visible, very fine noise that’s constantly buzzing around at all brightness levels. This very fine noise is easily mistaken for extra picture detail, which it is not.
It’s really interesting to look at side by side comparisons of the EX and XF footage. With the same clip, trees and foliage from the XF appear to be more detailed, but step through the footage frame by frame and you see the XF foliage is actually full of this fine noise compared to the EX foliage, this makes it appear sharper as it adds a fake “texture” to the foliage. But I can’t actually see any extra real detail in the XF foliage. Taking the same clip and looking at the parked car, again the XF shows a lot of noise, but in this case you can clearly see the car is a little softer in the XF footage than the EX. These differences are not so much down to resolution differences, but down to noise and noise reduction working in slightly different ways in the two cameras plus differences in the detail settings.
Turn up the gain on both cameras to +12db and the difference is even more striking. Click on the image to the left to view it full frame or look at the clip I have prepared (Vimeo or YouTube). I find the fine, busy, noise from the XF305 a lot more objectionable than the more blocky noise that the EX1R generates. The XF305 also shows some black speckles similar to those found on the Panasonic HPX301. Given that the XF305 is using small 1/3? sensors this kind of performance is not really unexpected. The XF305 probably has the best front end of any 1/3? camera currently on the market, but controlling noise on small sensors is harder to do than with large sensors and IMHO it still doesn’t perform as well as the EX’s with their larger sensors.
In addition the EX1R’s larger sensor helps it capture more light making it 1.5 stops more sensitive than the XF305. So in low light you will tend to use more gain with the 305 than with an EX. It has been pointed out that the XF305 lens is about half a stop faster than the EX lens which does help the 305 little, but if low light performance is important to you do choose wisely. You can increase the noise reduction on the XF305 to combat the noise but this also softens the picture, especially when used at the higher settings. If you download the raw footage you can zoom in and see the differences for yourself.
The differences are small, but they are there. The Canon is noisier looking while the Sony appears a little softer, but I’m not convinced that it is, I thinks it the XF’s noise giving the impression of a sharper image by adding texture to many surfaces. I personally would take the cleaner image as you can more with this in post production.
I also shot some clips with the XF305 and a NanoFlash as well as the EX1R and a NanoFlash. At 50Mb/s it was very hard to tell any difference between the XF305 and 305/NanoFlash recording, which is what you would expect. If anything the NanoFlash footage may be just a tiny bit less blocky. Comparing the XF305 at 50Mb/s and EX1R/NanoFlash at 50Mb/s the EX footage was quite a bit cleaner with less mosquito noise and macro blocking. If you click on the side by side image on the right you can view it full screen. Look at how much cleaner the EX footage is, but also look carefully at the brickwork. I can tell you that in the moving video clip the bushes in the XF305 clip are full of fine noise, on the frame grab this looks like fine detail, but it’s not.
The EX footage recorded on to the NanoFlash at 50Mb/s is cleaner with less macro blocking and mosquito noise, this is probably due to the very fine noise from the 305 stressing the codec harder than the low frequency noise from the EX. At 100Mb/s the EX1R looked really good indeed. The test footage was shot at 25P and even though the XF305 is recording at 50Mb/s 4:2:2 against the EX1R’s 35Mb/s 4:2:0 the visible difference in the rushes is negligible.
The rear viewfinder on the XF305 is a bit better than the rear finder on the EX1R. It’s bigger and I found it very nice to use, however I could not get the waveform monitors etc to appear in the rear finder, only the LCD screen. The menus are logically laid out, I didn’t find them as straight forward to use as the EX menu’s but then I have been using EX’s for some years now so it could just be a case of them being different to what I’m used to.
CONCLUSIONS
So what do I think? Well the XF305 is a very good camcorder. It produces very pleasing images recorded at 50Mb/s 4:2:2. Is it better than an EX1R? Well I don’t think that overall it is. Is the EX1R better then? On it’s own, no, but with a NanoFlash, yes. There are bits of both cameras that I like and dislike. The EX1R is nicer to hold and more compact, it has a better zoom rocker and record switch. The XF305 has a better zoom range and less CA (Chromatic aberration) but I would not enjoy the way you have to choose between the zoom rocker OR zoom ring and can’t have both. The 305?s zoom and iris rings are a little sluggish to respond and the iris has no calibrated markings so you have to rely on the viewfinder or LCD. The 305?s LCD is really very good and I like the way you can flip it out either side. The EX1R’s rear finder, while perfectly useable is not as good as the one on the XF305. The picture quality from both cameras is very good. Different…. but good. The 305 has a slight edge on it’s out of the box look but it is visibly more noisy than the EX1R and it’s easy to confuse the very busy fine noise that appears across the whole image as fine detail. The EX can be dialled in to give great pictures too. As a side note with the EX going in to the Picture Profiles and increasing frequency to +40 helps sharpen up the foliage in the EX pictures. Both cameras have some noise in the images, at 0db I think the EX has the visual edge and looks cleaner, at higher gain levels above +6db the XF305 noise becomes more and more objectionable compared to the EX1R. At 0db the XF305?s fine noise is stressing the codec a little. An EX1R recording to a NanoFlash at 50Mb/s produces a much cleaner image with less mosquito noise and macro blocking.
If your thinking of buying either it’s a tough choice. The XF305 has genlock and timecode in, which the EX1R does not have. For that you need to get an EX3 which is more expensive, but then you can also change the lens. I certainly don’t see any reason to swap my EX’s for 305?s and the similarly priced EX1/NanoFlash combo is an extremely powerful tool offering the benefits of dual record, HD and SD recording as well as higher bit rates. In addition NanoFlash 50Mb/s files are compatible with the XDCAM HD optical disc system, which the Canon files are not. The XF305 has smaller sensors than the EX1 so controlling depth of field will be a little harder, also you will get image softening due to diffraction effects about a stop sooner with the 305 but this may or may not be important to you.
In Summary:
Optics: Canon
Ergonomics: Sony
Pictures: Sony (cleaner, less noise)
Workflow: Sony (Because it’s faster (with SxS) more mature and you have backwards compatibility with Optical Disc XDCAM HD)
Canon XF305 and Sony EX1R side by side tests from Ingenious TV on Vimeo.
Alister,
I sent you an e-mail, but thought I would comment here too. I like all the features of the Canon XF 300 and 305 over the EX1R but will need good low light performance at times. I have been reading about and seeing examples of some post production NR plug ins (such as Neat Video) that look good. Is it possible to crank up the Canon’s gain to 18 and make it look like the EX1R looks at 12 by using careful NR in post-production? In other words, is there a way to make the Canon look acceptable for moderate quality TV broadcast work when shot at 18 db of gain? Thanks.
The XF305 is a good camera, but it has a small sensor and that means small pixels, so low light performance will not be as good as a comparable camera with larger pixels. Talking to various producers last week at BVE using the XF305, the general consensus was that in low light it was problematic with noise becoming an issue quite quickly. Both the 305 and Sony’s EX cameras already use internal noise reduction circuits, so further noise reduction via software will be limited in scope without reducing image quality. To be honest using any of the small chip camcorders with more than 6db gain is unlikely to be acceptable for broadcast work unless there is strong justification for it. I would certainly never consider using 18db of gain on any kind of broadcast production, the only exception being the new Sony F3 which at +18db gain has no more noise than an EX1 at +6db.
Thanks Alister. I ended up buying the PMW-350k. I just got tired of trying to figure out how to make a smaller camera produce the same quality image. Part of the reason I bought it was because of your review, but there have also been other good reviews of that product. I will likely add a Nanoflash to it eventually. Do you know if there is a good aftermarket bracket/holder for the Nanoflash for that camera? Thanks for all you do to help folks like me.
@alisterchapman
You can’t judge cameras purely on sensor size as resent development in sensor technology has changed the ball game somewhat. I own a couple of XF305’s and within the last six months a C300 – this is personal kit, not hired.
If you’re prepared to dig into the XF305’s possibilities, you’ll discover a camera that offers outstanding image quality.
The only thing it lacks is extreme shallow DOF … I can grade the XF305 & C300 into any edit seamlessly – so that ‘cine’ look is more than possible. The only negative is the XF305 is very poor in auto mode – keep everything manual and set CP to Cine F (near flat) and post grade.
I toyed with a loaned 1Dc and filmed at 4k … even this DSLR look-a-like gave me a glimpse of the future – we’re talking RED quality at a fraction of the price. Sensors are important, but the size of a sensor isn’t as important as it used to be.
The 305’s are a valued tools in my kit though the C300 is better overall it dies cost considerable more.
Oh … and as for the EX1 … I owned this beauty for three years, but I don’t feel as if I’ve lost anything – the move I’ve made is an upward one.
There really have not been any changes in sensor technology for some years and there are unlikely to be any big changes any time soon. Modern sensors all rely on silicon and silicon has a limited efficiency at turning photons of light into electrons of electricity. Most modern sensors achieve efficiencies (QE) of around 70%. There are some developments occurring in the use of colour splitters instead of colour filters for bayer type sensors, but really sensor technology has remained largely static for several years. Sensor and pixel size is the single most important thing in the front end of a video camera because it determines everything else that follows.
Video sensor performance is governed primarily by pixel size. This is simple laws of physics stuff. The bigger you make the pixel, the larger it’s surface area, the more light it will capture. In addition larger pixels can hold more electrons before they overflow, so dynamic range also increases.
What has changed and improved is noise reduction. Noise reduction techniques have improved and the amount of processing power available for noise reduction has increased. But noise reduction is not perfect, it always comes at a price, this can manifest itself in the image quality in different ways. Image smear, softness, ghosting, banding, etc are typical NR artefacts. At lower NR levels these may be small enough to go undetected, but as noise levels increase, for example when adding gain they become more pronounced. The XF305 is a perfect example of this. At the nominal 0db point the XF305 and Sony EX1 appear to perform similarly. But start adding in gain and the XF305 image degrades much more rapidly. This is because even at the nominal 0db point the 305 is already using more NR than the EX1 to compensate for it’s lower base sensitivity. Adding gain is pushing the NR harder with the 305 than EX1, so it falls apart faster when the light levels fall away. This is the most common complaint I hear about the 305, it’s poor in low light. That is due to the smaller sensor and thus pixel size.
If you really think that the 1Dc with it’s 8bit jpeg encoded 4k is comparable to a Red camera shooting raw then I would suggest you look again at what is possible with raw footage compared to 8 bit jpeg, the difference is night and day. The 1Dc is a very interesting camera and it produces a great image. I would consider one for covert 4K shooting or as a grab and go 4K camera but it is not in the same league as Red.