Hello all, I am excited to share this test report I wanted to make ever since I acquired both lenses, in an effort to help the community.
I will also attach it to the thread for easy download. Enjoy your read and I hope this comparison helps you make a purchase.
Astrophotography Report
Canon EF 300mm f/4 L (1991-1997) vs Canon EF 300mm f/4 L IS (1997-2022)
1. Introduction
It is commonly assumed that newer optical designs are inherently superior to older ones. In the vast majority
of cases this assumption holds true. However, there are notable exceptions, particularly among high-end
telephoto lenses designed in the early 1990s.
The Canon EF 300mm f/4 L USM, released in 1991, is one such exception. The lens was discontinued after a
relatively short production run and replaced by the Canon EF 300mm f/4 L IS USM, which introduced image
stabilization and a substantially revised optical formula. As a result, original non-IS copies are comparatively
rare on the used market.
I initially acquired the IS version, with the expectation that its optical performance would at least match, if not
exceed, that of its predecessor. While I will not discuss optical design schematics or glass layouts, as end
performance is what ultimately matters in astrophotography, it is important to note that the two lenses differ
significantly in construction.
A recurring claim in online discussion is that the IS version is optically inferior to the original, particularly in
terms of sharpness. Based on my testing, this claim is broadly correct. However, sharpness alone is not the
primary reason why I consider the stabilized version less suitable for astro imaging.
All observations and conclusions presented in this paper are derived from direct use and personal visual
analysis. Perception of optical performance can vary depending on individual preferences and imaging
priorities, especially in astrophotography, where star morphology and contrast are often valued differently than
daylight shooting.
2. Testing rigor and methodology
Equipment used:
Canon 6Da (both) + Bahtinov mask
Sky-Watcher Star Adventurer Pro (300mm IS)
Fornax Mounts Lightrack II (300mm non-IS)
We are using a relatively forgiving sensor, which can mask resolution limits but does not hide low-order
aberrations such as coma, haloing, or deformations.
It has already been established that comparing single exposures is meaningless as no significant differences
can be detected. The followed testing protocol is a very loose “apples-to-oranges” comparison. External factors
introduce variability in testing, such as:
Integration times are different – analysis on “stacked data” but no integration time match;
Temperature conditions vary from night to night, and so do humidity and seeing;
Tracking accuracy will be different. The Star Adventurer Mount the IS-version was tested on presents
a higher degree of periodic error, influencing oblongness of the stars and their full-width-halfmaximum,
thus FWHM values will not be analyzed for sharpness purposes;
Different targets and/or orientation;
Focus accuracy can vary.
Regardless of the aforementioned issues, the following analysis should still provide an insightful view of the
optical difference between the two lenses. Sample variation is also a factor which cannot be controlled or
considered; however, Canon lenses are notoriously good at varying by very little among copies. The stabilized
version, however, is shown to vary more than its older counterpart.
Link to the most coherent analysis for objective MTF curves measurements between the two lenses:
https://www.wlcastleman.com/equip/reviews/300mm/index.htm
3. Optical performance
Field correction, distortion and coma
Both lenses perform well in terms of overall field correction. On a full-frame sensor, corner performance is
generally strong, with a slight advantage observed in the older non-IS version.
The IS version exhibits a degree of uneven coma at f/4, which varies in direction depending on the corner of
the frame being analyzed. This behavior happens to be sample-dependent, but this was consistently present in
my copy. The effect is largely mitigated by stopping down the lens.
Overall, distortion is negligible in both lenses, which is typical for super-telephoto designs and largely
irrelevant for astrophotography use.
Remember that:
𝑠𝑡𝑒𝑙𝑙𝑎𝑟 𝑓𝑖𝑒𝑙𝑑 𝑎𝑏𝑒𝑟𝑟𝑎𝑡𝑖𝑜𝑛𝑠 ≠ 𝑙𝑒𝑛𝑠 𝑑𝑖𝑠𝑡𝑜𝑟𝑡𝑖𝑜𝑛
Haloing and stellar core contrast
This category is separated from conventional sharpness analysis, as the two are not equivalent in the context
of astro imaging. Certain lenses can appear exceptionally sharp when tested on high contrast terrestrial targets
yet perform poorly when imaging point light sources such as stars.
Complex optical systems can produce halos, scattering, or asymmetric star artifacts depending on how light
interacts with internal elements. For astrophotography, these effects are more critical than resolving power by
itself.
non-IS version
The non-IS lens produces very light contrast loss at f/4 but this is far from a bad performance.
That mixed with some blue fringing makes the stars a bit bigger than what you would find in an APO triplet.
IS version
The stabilized version displays more problematic behavior. After stretching, multiple triangular halos are
visible around stars, with a clear tendency to orient toward the center of the image frame. This effect is not
limited to extremely bright stars (such as Alnitak in Orion) but is also clearly visible on moderately bright
stars.
The severity and orientation of these artifacts appear to be sample-dependent. In my case, the top-left quadrant
of the frame was most affected. Additionally, bright stars exhibit low-contrast, asymmetric smearing, which is
noticeable, though, not catastrophic.
Stopping down the lens to approximately f/4.5 using mechanical step-down rings almost completely eliminates
the asymmetric smearing. However, the triangular halos persist even after stopping down. For this reason, the
use of step-down rings is recommended for both lenses, though it does not fully resolve the IS’s haloing
behavior, shown in the file.
One solution would be to take exposures of the stars separately and avoid extreme stretching when blending
them into a starless image. Although not ideal, it is a profitable workaround.
Pinched optics
Both lenses show no visible signs of pinched optics under
typical operating conditions.
The non-IS occasionally exhibits a two-lobed or “butterfly-like” star shape
on bright stars, which could be indicative of aperture vignetting effects. This
effect is subtle and does not meaningfully degrade image quality. In some
cases, it may even be aesthetically pleasing.
Neither lens shows behavior consistent with severe pinching, and both remain far from scenarios where image
quality would be considered compromised.
Chromatic aberration
A slight advantage can be attributed to the IS version in terms of CA control. With built-in corrections enabled,
chromatic aberration is very well controlled and only a faint trace of magenta fringing becomes visible after
aggressive stretching of approximately one hour of stacked signal.
The non-IS lens performs slightly worse in this regard, though its aberrations are largely confined to the blue
end of the spectrum. For astrophotography, this is often less objectionable, as blue fringing typically manifests
on blue stars, where they are, of course, blue, thus remaining consistent with stellar color.
No red fringing was detected at infinity focus on the non–IS lens.
When built-in corrections are disabled, both lenses perform similarly, and no definitive winner can be
established.
Actual transmission
Despite vignetting correction being applied, the IS version appears to transmit approximately one-third of a
stop less light when used wide open.
Observed settings (IS version)
f/4 – ISO 1600 – 1-second histogram peak slightly left of center
f/5.6 – ISO 1600 – 2-seconds histogram peak centered
When vignetting correction is applied to the non-IS lens (which lacks built-in profiles), a similar but less
pronounced effect is observed. In this case, the light loss can largely be attributed to off-axis ray falloff, which
is normal for fast telephoto lenses.
Overall, the IS lens should be expected to be slightly slower in effective transmission. This could also be due
to the higher glass density. The difference does not significantly influence the final recommendation.
Actual focal length
Plate solving reports the following effective focal lengths:
294mm for the IS;
296mm for the non-IS.
This is on par with the manufacturer’s claim. It can slightly vary with sample variation. Effective focal length
combined with entrance pupil diameter determines the true f-number:
294mm/75mm
= 𝑓/3.92 , 𝑛𝑜𝑡 𝑓/4
Sharpness
IS version
At f/4, the IS version appears noticeably muddier, with reduced microcontrast. By f/5.6, sharpness improves
dramatically and becomes excellent all around.
When used with 62mm step-down rings, the lens delivers very good sharpness and an effectively flat field.
This configuration represents the optimal operating point for this optic. However, it does not eliminate the
previously discussed haloing artifacts.
non-IS version
The non-IS lens is exceptionally sharp even at f/4, with corner performance becoming excellent at f/5.6.
Stopping down to approximately f/4.5 using a 67mm step-down ring yields good improvements for star
rendering. This setting represents the optimal balance between sharpness and throughput for astrophotography
use.
Daylight photography
In daylight use, the two lenses are difficult to distinguish in terms of central sharpness. Differences are most
prevalent in the corners, where the non-IS version is noticeably sharper.
For multipurpose use, the presence of image stabilization makes the IS version more practical for daytime
photography, and this tradeoff may be worthwhile depending on your use case.
4. Mechanical usability & budget considerations
It is often argued, correctly, that dedicated telescopes outperform camera lenses for astrophotography.
However, budget considerations introduce an important caveat.
The non-IS lens evaluated here was purchased for €309 in non-functional autofocus condition. At this price
point, it competes with very few optically corrected, flat-field systems capable of covering a full-frame sensor
and of shooting at a fast f-ratio.
The closest well-corrected competitor in terms of price, focal length, and speed, is the SvBony SV555,
which costs €599 as of February 2026. While it offers good optical performance and useful accessories,
its built quality is significantly inferior to that of the Canon lens, which is borderline indestructible. Similar
concerns apply to the Askar SQA55, which, despite excellent optical quality, costs approximately €980 and is
reportedly subject to poorer build quality, which goes a long way for durability.
In my assessment, a well-corrected FPL-51 doublet does not outperform the non-IS canon lens when the latter
is used at f/5. Its first true competitor, in my view, is the Askar FRA300 Pro, which offers excellent optical
quality, solid construction and a fast focal ratio at a 300mm focal length. However, at approximately €1100, it
costs roughly three times as much as a typical used Canon EF 300mm f/4 L.
Dedicated telescopes remain superior in terms of usability. They are designed for infinity focus, feature precise
focusing mechanisms, built-in guiding solutions, filter drawers, and numerous quality-of-life improvements.
Nevertheless, on a limited budget, a high performing telephoto lens such as this remains a compelling option.
Many users continue to employ this optic with dedicated cooled astronomy cameras and custom EAF systems,
finding no compelling reason to upgrade. This is a testament to how good this lens actually is, given you
provide some additional care to it.
5. Conclusions
The Canon EF 300mm f/4 L IS USM is not recommended for astrophotography use due to persistent triangular
halo artifacts and reduced performance at wide apertures.
The older Canon EF 300mm f/4 L USM, by contrast, is strongly recommended as a budget-friendly alternative
to a well-corrected apochromatic refractor. It exhibits very good sharpness wide open, low coma, and, most
importantly, does not produce intrusive halos.
Both lenses will require careful focusing, but hourly refocusing with a Bahtinov mask is manageable in
practice.
The haloing behavior of the IS version proved to be a decisive limitation of my use case. I ultimately sold the
lens at approximately the same price at which it was purchased, resulting in a happier wildlife photographer.
The lens is likely better suited for whom its image stabilization provides a meaningful use.